The Structural Anatomy of the Norwegian State 1947-2003

(1)

The Structural Anatomy of the Norwegian State 1947–2003

PE R LÆ G R E I D

VI D A R W. RO L L A N D

PA U L G . RO N E S S

JO H N- ER I K ÅG O T N E S

ST E I N RO K K A N CE N T R E F O R SO C I A L ST U D I E S U N I F O B A S

D E C E M B E R 2 0 0 3

Working Paper 21 - 2003

(2)

Preface

¹

This paper is written in connection to the research project on «Regulation, Control and Auditing» by members of the research group on administration and governance at the Rokkan Centre. The project is funded by the Norwegian Research Council. Earlier versions of this paper has been presented at the seminar on the study of public sector organization in Leuven, May 2–3 2003 and at the seminar on organizational forms, autonomy and control in the public sector in Bergen, December 1–2 2003. We wish to than the participants at the seminars for valuable comments.

1 Paper prepared for seminar on organizational forms, autonomy and control in the public sector , Bergen, 1–2 December 2003.

(4)

Summary

This paper describes and analyses the development of the Norwegian state apparatus in the post World War II period based on a broad institutional perspective and a comprehensive database. The Norwegian reform context is presented by focusing on the main components of the administrative policy that different governments have pursued, and the main organizational forms in the Norwegian state is described. The empirical data is presented by focusing on changes in formal organizations over time:

the size of the state apparatus, and horizontal and vertical specialization. A main empirical finding is that there has been an increasing parallel process of vertical specialization and horizontal de-specialization in the Norwegian state over the past 20 years.

(5)

Sammendrag

Dette notatet beskriver og analyserer utviklingen i den formelle organisasjonsstrukturen i den norske stat i tiden etter den andre verdenskrig, basert på et bredt institusjonelt perspektiv og en omfattende databank om strukturelle endringer i norske statlige virksomheter i perioden 1947–2003. Den norske reformkonteksten blir presenteret ved å fokusere på hovedkomponentene til de ulike regjeringers forvaltningspolitikken i denne perioden. Det gjøres videre rede for de viktigste formelle organisasjonsformene i staten. Det empiriske datamaterialet blir presentert ved å fokusere på endringer i formell organisering over tid, på endringer i størrelsen på statsapparatet og på horisontal og vertikal spesialisering. Et hovedfunn er at det i løpet av de siste 20 årene har vært en økende tendens til en parallell utvikling i retning av sterkere vertikal spesialisering samtidig som det har foregått en sterkere horisontal de-spesialisering. Dette innebærer at virksomheter har blitt fristilt ved at de har blitt flyttet fra tilknytningsformer som ligger nær politisk ledelse til tilknytningsformer som ligger på større avstand samtidig som enheter på samme forvaltningsnivå slås sammen til større enheter.

(6)

(7)

Introduction

When discussing the reorganization of public sector organizations it is important to distinguish between reform and change. By reform we mean the active and conscious attempt to change the organizational structures, procedures or personnel in administration. Based on this definition, we can imagine situations where reforms do not result in changes, or situations where changes occur that are not the result of reforms (Brunsson and Olsen 1993). In this paper we are concentrating mainly on structural changes and ask to what degree the changes can be understood as a result of an active administrative policy, or if we have to take other considerations into account to understand the changes.

The state apparatus is a multi-functional system, where different units may not merely maximize a simple set of goals but are required to compromise between partly conflicting objectives and values. Ambiguous and inconsistent goals are inherent characteristics of the public sector, which political and administrative leaders have to find ways to live with, rather than a sign of disease that needs to be cured. This implies that political and administrative leaders are permanently confronted by tensions and dilemmas to which there is no simple solution (Lægreid and Roness 1999). Changing the organizational form of state organizations may challenge and change the balance between different values and tasks, like political control and professional or corporate autonomy.

The structural development of the state administration can be described along a combination of a vertical and a horizontal dimension (Egeberg 1989, Christensen and Egeberg 1997, Lægreid and Roness 1998). By vertical specialization we mean differentiation of responsibility on hierarchical levels, describing how political and administrative tasks and authority are allocated between forms of affiliation. Vertical specialization can take the form of structural devolution, autonomization or agencification, meaning transfer of responsibility from units close to the political leadership to units that are further away from the political executives. Vertical de- specialization implies movement in the opposite direction by moving responsibilities closer to the political leadership. By horizontal specialization we mean splitting of organizations at the same administrative level, for example splitting a ministry (or an agency) into several ministries (or agencies). Horizontal specialization focuses on how tasks and authorities are allocated between organizations at the same hierarchical level,

(8)

for example between ministerial areas (Christensen and Peters 1999). Horizontal de- specialization implies merging of organizations at the same administrative level.²

As shown in Figure 1 there can be quite complex change processes going on in individual reform cases, involving both horizontal and vertical specialization or de- specialization. As we will show later, the most common form of change in the Norwegian state apparatus since 1947 is horizontal de-specialization without changes on the vertical dimension. Less common, especially in the last decades, is vertical de- specialization. In this paper we will handle the vertical and the horizontal dimension separately, and not dig into the complexity of combining them.

Figure 1. Horizontal and vertical specialization and de-specialization

HORIZONTAL SPECIALIZATION

Recoupling/no change HORIZONTAL DE- SPECIALIZATION

VERTICAL

SPECIALIZATION Example: Splitting the integrated organization Norwegian Railways into a company (NSB), a regulatory agency and a directorate

Example:

Transforming the Norwegian Broadcasting Corporation from a public administration body into a state- owned company

Example: Establishing the central agency for health and social services by moving tasks from ministry and merging agencies

Recoupling/no change Example: Splitting the Ministry of Social Services and Health into two ministries

Stability/internal

changes Example: Merging of colleges

VERTICAL DE-

SPECIALIZATION Example: Integrating

the Headquarter of Defence and the Ministry of Defence

The paper describes and analyses the development of the Norwegian state apparatus in the post World War II period. Three questions are topical:

1. What are the changes in the size and anatomy of the state apparatus in this period?

2. What is the degree of horizontal and vertical specialization in the state apparatus in this period?

2 In addition, we can distinguish between intra-organizational and inter-organizational changes in structural arrangements (Egeberg 1989, Christensen and Egeberg 1997). By this we mean if the specialization occurs within or between organizations in the state administration. Intra-organizational changes are changes within distinct organizations such as a ministry, an agency or a state-owned company. In this paper we will mainly focus on inter- organizational changes.

(9)

3. How can we explain the changes that have occurred in the structural arrangements? To what extent are they results of an active administrative policy, and to what extent are they constrained by internal administrative culture and tradition or by external contexts, situations and the dominating international doctrines of different periods?

We will first outline the database and the method used. Second, we will present our theoretical approach, which is a broad institutional perspective. Third, we will present the Norwegian reform context by focusing on the main components of the administrative policy that different governments have pursued. Fourth, we describe the formal structural arrangements in the Norwegian state by focusing on the main organizational forms. Fifth, we will present our empirical data in two sections focusing on changes in formal organizations over time: the size of the state apparatus, and horizontal and vertical specialization. The paper ends by discussing the coupling between administrative policy and changes in structural arrangements and concludes by pointing out the main findings.

Data and method

The empirical basis for the paper is a review of existing studies of structural devolution reforms and administrative policy in the Norwegian state apparatus, and a comprehensive Norwegian State Administration Database (NSA) covering changes in the formal structure of the Norwegian ministries, civil service, state-owned companies and governmental foundations from 1947 onwards. An organization is included in the database if it meets one of the four following criteria:

a) It is a ministry or subunit within ministries, like divisions, sections or offices.

b) It has a legal status as part of the state, separate from, but subordinated the parent ministry or parliament. These organizations are labelled civil service.

c) It is state-owned or partly (i.e. majority) state-owned. These are labelled state- owned companies (SOCs).

d) If a self-owned organisations is founded by ministry or central agencies. These are labelled governmental foundations.

Within the single group of state organizations there are some further sub-categories. For the organizations that are included in the NSA, changes in organizational structure from birth to death are recorded. A predefined categorization that handles organisational change in three main categories is used: changes related to the founding of an organisation, changes related to the maintenance of an organisation, and changes related to the ending of an organization. For each main category of changes there are several sub-categories, like secession, dividing, merger, absorption and movement of organizations vertically and horizontally within the state apparatus and out of it.

As of 2003, the database covers a total of about 2.850 unique organizations, including 1.700 internal units within the ministries. These 2.850 unique organisations

(10)

have been through about 8.600 organizational events, indicating that there are substantial turbulence and structural changes of different kind going on in the state apparatus.³

Theoretical approach

The discussion will primarily be based on a broad institutional perspective. An institutional perspective implies that, over time, organizations will develop an

‘institutional profile’, comprised of those roles, rules, regulations, routines, traditions, standard organizational forms and procedures which are the common features of institutions. We expect that these characteristics will provide constraints on how the organizations react to external pressure form the environment as well as to politically motivated reform initiatives. Thus, in order to understand reforms and changes in the administrative system, we will be concerned with the relations between the reform programmes, institutional constraints and societal changes. An institutional perspective would question that changes in organizational structure are a prompt and necessary response to changing demands from the environment as well as being an immediate and straightforward response to political signals (Olsen 1992, Lægreid and Roness 1998).

The chances for planned design are restricted, though not to such a degree that the political actors are rendered powerless.

Like other organizations, the state organizations may have institutional characteristics, including specific identities and capabilities (March and Olsen 1995, Roness 2001a). Some of the identities and capabilities are expressed through elements of specialization and hierarchy within and between organizations. At any given time, the historical legacy of an organization is ingrained in structural features that will delimit or open up for specific actions and outcomes. Structural change means new forms and degrees of specialization and hierarchy within and between organizations, implying changing identities and capabilities. Specialization is primarily related to the handling of certain policy areas or tasks in separate parts of the organization, in central government by the various ministries and agencies and in the state-owned companies and government foundations between different types of companies and foundations.

To understand the diversity of the structural devolution processes and their effects and implications we will use a transformative perspective that focuses on the complex and dynamic interplay between instrumental reform efforts, national structural and cultural contexts, organization-specific characteristics and external environmental pressures (Christensen and Lægreid 2001a, Olsen 1992).

3 For a more detailed description of the database see Rolland and Ågotnes (2003), and for a discussion of the problems of categorizing state organizations, see Roness (2003).

(11)

The national context identified by the transformative perspective concerns first the formal aspect of the political–administrative system and its implications for the instrumental reorganization of state organizations (Campbell and Halligan 1992, Christensen and Peters 1999). Norway has for several decades had minority governments, and over the past decade an increasingly active parliament. The degree of political or administrative control over state organizations versus professional and commercial autonomy is related to how this vertical relationship is formally organized.

This means that the traditionally strong political control of government bodies in a country like Norway, within a system of ministerial governance, goes hand in hand with a hierarchical and integrated model. From an instrumental perspective we will, first, expect a tight coupling between administrative policy programmes and structural changes in the state apparatus. The goals, values and suggested solutions that are enhanced in the central policy programmes will affect the changes in the different structural arrangements. Second, we will expect changes over time in organizational forms in accordance with the suggested reforms in the policy programmes.

The national context also concerns the historical and cultural traditions of political–

administrative systems (March and Olsen 1989, Selznick 1957). Informal norms and values develop over time, are manifest in a distinct administrative culture and serve to further stability and integration. The point of departure in this approach is that a certain style of controlling government agencies and state-owned companies has developed over the years, ranging from countries that have tended to see state-owned companies as a strong and integrated instrument of political development, to countries where leeway and autonomy for SOCs and their loose coupling to political actors or goals have been taken for granted (Grønlie 1998). The crucial question in a period of reform and transition is whether traditional norms and values associated with SOCs match the norms and values of the modern reforms (Brunsson and Olsen 1993). Clearly, a lack of compatibility may prevent or modify the reforms, while compatibility may reinforce them. Administrative reforms are more easily accepted if they are in accordance with the administrative culture. From a cultural point of view we will, first, expect less variation in structural arrangements over time. The adaptation process will be gradual, incremental and slow. Second, since there may also be some organization-specific cultures, we will expect variation in structural arrangements between ministerial areas.

Internal rules, traditions and routines will be different from one sector to another, and path dependencies will constrain how the formal organizations change in practice. The internal forces for change, robustness and stability will be more important than shifting policy signals, and the coupling between the general policy doctrines and actual changes will be rather loose.

A third set of factors describes the changes of formal structural arrangements as a response to external pressure (Olsen 1992). This environmental determinism can be of two kinds. In the first instance, a country may adopt internationally based norms and beliefs about how a public business should be organized and run simply because these have become the prevailing doctrine. New Public Management (NPM), with its strong focus on structural devolution, has its origins in certain Anglo-Saxon countries and international organizations, like the Organization for Economic Cooperation and Development (OECD), where a kind of reform myth has taken hold, becoming

(12)

ideologically dominant and diffused all over the world (Czarniawska and Sevón 1996, Sahlin-Andersson 2001). This diffusion process may imply isomorphic elements – i.e. it may create pressure for similar reforms and structural changes in many countries (cf.

DiMaggio and Powell 1983). Administrative policy rhetorics may reflect international trends and ‘fads’. The looser the association between formulations in reform programmes on the reform content and the structural changes in practice, the more the reform may be interpreted from a symbolic perspective. The ideas of institutional environment will predict variation in actual organizations over time. What the dominant doctrine of the day is will vary over time, and what is regarded as appropriate and modern organizational forms will influence the actual changes. There is a decontextualization process going on that will lead to small differences between ministerial areas.

In the second instance, structural devolution may be adopted to solve widespread problems created by a lack of instrumental performance, by technological change or by economic competition and market pressure in a globalized economy. In this case, reforms of state-owned companies are adopted not because of their ideological hegemony but because of their technical efficiency. The absence in Norway of national fiscal pressure and the accumulation of a huge governmental petroleum fund may reduce this kind of external pressure for devolution. On the other hand, the external constraints coming form increased integration into the Europe through the agreement of the European Economic Area (EEA) that came into force in 1994, might lead to isomorphism between the Norwegian structural arrangements and the EU doctrines.

The EU policy towards market competition and deregulation are at the heart of New Public Management. The EU puts normative pressure on the domestic administration, and also more directly demands certain ways of organizing and controlling state apparatus to secure competition in the internal market, like the establishment of autonomous control and regulatory agencies (Bleiklie, Lægreid and Wik 2003, Jacobsson, Lægreid and Pedersen 2003).

We will argue that public reform processes and their effects on organizational structures are not characterized by a simple adjustment to current international administrative doctrines, but must be understood from a transformative perspective. Taken together, the environmental, cultural and structural contexts constitute transformative preconditions and constraints that in a complex and dynamic way define the leeway political and administrative leaders have in responding to global reform ideas and to specific reform initiatives (Christensen and Lægreid 2001a). Cultural factors can modify the effects of administrative policy programmes on the reorganization of state organizations. But cultural factors can also support the structural devolution, making the effects even stronger. In the same way, external pressure exerted by new administrative doctrines might be enhanced or tempered by domestic political strategies or by an administrative culture or tradition. The transformative perspective adds complexity to our understanding of administrative reform processes and makes the story less elegant but probably more realistic.

(13)

The Norwegian reform context – administrative arguments and doctrines

Administrative reforms are not new phenomena in the development of the Norwegian state. What is new, however, is that starting in the early 1980s, the field of administration emerged as an independent policy area with its own vocabulary and institutions, in which reform initiatives were taken up through comprehensive and universal reform programmes. The central administrative policy documents include the modernization programme of the Willoch government (1986), the renewal programme of the Brundtland government (1987), the report from the Hermansen commission on a better organized state (1989), the Jagland government’s «The Norwegian House» (1996), the programme on a «Simplified Norway» of the Bondevik I government (1999), the Stoltenberg government programme for innovation of the public sector in Norway (2000), and the programme of the Bondevik II government labelled «From words to action» (2002).

The Norwegian administrative system is a complex body combining partly conflicting principles, values and mixtures of different control devices. Most of the basic values and considerations in administrative policy have been present through the post- Second World War period, but the individual components and the emphasis attached to each have changed over time (Lægreid and Rolland 1994). During the first decades after 1945, a form of problem solving was undertaken through the establishment of new administrative units or reorganization of old units (Lægreid and Roness 1983). As soon as problems in society had been accepted as the responsibility of public sector, a permanent administrative body was established for furthering routine problem solving in that policy area. The result is a state apparatus that can be characterized as a coral reef, comprised of layer upon layer of organizations established at various points in time, without any collective overall plan. As we shall see, however, the permanence of these organizations is not as high as the coral reef metaphor indicates. They are more or less permanently under reorganization through splitting, merger and movement vertically and horizontally in the state apparatus^.

There appears to be a cyclical process, whereby development in one direction results in a counter-process and sometimes the re-emergence of new values and controls in another. In the 1950s and 1960s, considerable importance was attached to constitutional law and collective standards for appropriateness and justice. The rule of law, hierarchical structure, formal control and traditional bureaucratic procedural methods of administrative control was underlined. The Parliamentary Ombudsman for public administration was established in 1962, and a Public Administration Act came in 1967.

A main question in the Norwegian administrative history from the mid-19th century has been the dispute on internal structural devolution, meaning if the central administrative bodies should be organized in the form of ministries or directorates (central agencies). How much autonomy directorates should have in relation to ministries has been a recurrent problem. Form the mid 1950s, a dominant

(14)

administrative doctrine was that the ministries should be relieved of routine tasks, which were administrative and technical in nature, and that these should then be transferred to subordinate directorates and agencies. Another important doctrine was that the ministries should be developed into secretariats for political leadership.

From the early 1970s, there was a stronger focus on political decentralization to counties and municipalities at the expenses of the central agencies. A new political body at the regional level was established, and the policy programmes from the mid 1970s announced that tasks and responsibilities should be moved from ministries and central agencies to this political–administrative level.

The 1970s was a period where importance was attached to increased transparency, participation and representation through the development of internal workplace democracy in public administration and by transferring tasks and responsibilities to a new regional level. The principle of free access to official records was introduced through the Freedom of Information Act in 1970. During this period, the aim was at political decentralization, rather than strengthening of the directorates. County and municipal authorities received responsibilities for a number of tasks, which were transferred from central government.

Norway has a long tradition of a homogenous and parliamentary-based political leadership living in peaceful coexistence. The political leadership maintained a close connection with administrative leadership characterized by strong mutual trust.

Traditionally, central political and administrative actors have agreed on balancing political considerations with the value of a rule-oriented civil service, citizen’s rights, transparency, equity, interest mediation and codes of professional behaviour. There has traditionally been a strong political control of public enterprises in Norway within a system of ministerial governance, and this went hand in hand with hiearchical structural features and an integrated administrative model. Up to the mid 1990s, Norway has tended to see public enterprises as a strong and integrated instrument of political development, organized as government administrative enterprises (Grønlie 1998).

The Hermansen commission (1989) changed the focus from internal structural devolution and the relations between ministries and directorates to external structural devolution focusing on state-owned companies and governmental foundations. The commission discussed the coupling between tasks and organizational form, and suggested more active use of different standardized forms of state-owned companies.

Governmental foundations were not recommended because of accountability and steering problems connected to this form of affiliation.

In Norway, the control function was for a long time, particularly after World War II, general and passive, allowing the executive a lot of leeway in general decision making, and more specifically letting the organizing of central government be an executive prerogative (Christensen and Peters 1999, Roness 2001a, Smith 1997). This seems to reflect some major features of the political–administrative system: a high level of mutual trust between political and administrative leaders and within the public administration, strong egalitarian values and common attitudes and norms among politicians and civil servants, clear role allocation between the powers and delegation of authority from the legislature to the executive (Christensen and Peters 1999, Christensen and Lægreid 2001a).

(15)

Norway has been characterized by a strong statist tradition, incremental changes, the balancing of many values in civil service and a policy style of peaceful coexistence and revolution in slow motion (Olsen, Roness and Sætren 1982). The relations between politicians and administrative leaders, and also within the public sector, have traditionally been one of cooperation and mutual understanding. Norway has thus a consensus-oriented political–administrative system. With its multi-party system and minority governments, it is generally characterized by negotiation and compromise in public policy-making processes. Adding to this, the civil service in Norway is, relatively speaking, more homogeneous in structure and personnel than the civil service in larger countries where the diversity is substantial on both counts.

Norway has, since World War II, experienced many parliamentary constellations.

From 1946 to 1961, the Labour Party had the majority and formed the cabinet alone.

Although it lost the majority in 1961, it remained in power until 1965 (except for a short time in 1963). Then came a period of non-socialist majority government, while the last three decades have mainly seen minority governments, either non-socialist coalitions or the Labour Party alone.

Considering that Norway belongs to a Scandinavian administrative tradition emphasizing the political–democratic context of the civil service, Weberian values, corporatism and equality, and a large universal welfare state, one may expect that the public administration will not adapt fully to NPM reforms.

The modernization programmes from the mid 1980s emphasized that the public sector had to make better use of resources and increase efficiency. Proposals were made for less control and greater autonomy. However, the reform programmes were a collection of reform ideas rather than a consistent policy. Opposing objectives such as increased autonomy, discretion and flexibility on the one hand, and enhanced political and administrative control on the other, were hardly discussed. The programmes had initially a strong management approach using management-by-objectives-technique and structural devolution, and the slogan ‘letting the manager manage’ was popular. Later on, control components by the use of performance management have been emphasized more strongly, and the use of contracts, marketization and outsourcing has become a major strategy for the government of today.

Discussion of NPM reforms began in Norway in the mid 1980s, but the Norwegian reform programmes then were more a loose collection of on-going reform measures and new reform ideas than a consistent, coordinated and unified strategic plan for changing the state apparatus (Lægreid and Roness 2003). The rhetoric in the modernization programmes reflected international trends in administrative policy, but it was not very specific concerning reform measures. The reform programmes promised better service, better utilization of resources, higher productivity, stronger attainment of goals, improved quality, better working conditions for employees, improved political control and extended user participation. Even though a vision of market- and management-based practices was presented in administrative policy documents, its implementation remained optional (Christensen and Lægreid 1998a, Naschold 1996: 66).

The assumption was that the pursuit of economic performance should not affect negatively other values like democratic representation, a state based on the rule of law and professional quality. Concerns about the implications of the reforms for democracy

(16)

and inequality modified the reform efforts (Lægreid 2001a), but in practice there has been a change in the balance of different values in favour of efficiency.

The reform programmes of various Norwegian governments thus exhibited more similarities than differences. They were pragmatic and cooperative rather than ideological and confrontational, and they were directed more at enhancing internal productivity and increasing efficiency than at rolling back the state. Reform strategies tended to be sector-based rather than comprehensive. The reform style up to the mid 1990s was maintenance, adjustment and supplementation rather than radical change.

Some elements, like the management components of NPM, have been implemented to a greater degree than the market elements (Christensen and Lægreid 1998a).

The Norwegian adoption of NPM through an active administrative policy under the auspices of the executive has since the mid 1990s mainly taken the form of increased structural devolution and the introduction of Management-by-objectives-and-results (MBOR) (Christensen and Lægreid 2001a). MBOR entails more flexibility, leeway, autonomy and discretionary power for subordinate agencies. It is a form of procedural autonomization or devolution, implying that the formal organizational structure is stable, but the internal rules and procedures are more flexible, opening up for more leeway for agency managers. Both structural devolution and MBOR illustrate the hybrid character in NPM coming from the centralizing tendencies in contractualism and the decentralizing tendencies in managerialism. NPM is thus a double-edged sword prescribing centralization, regulation and control as well as decentralization, flexibility and autonomy. Units and tasks are moved to organizational forms that are further away from the political leadership, implying more freedom for devolved agencies and generally weaker potential for control and scrutiny by the government, but also the need for new forms of control and regulation by use of contract steering and performance budgeting (Christensen and Lægreid 2001b). The role of the ministries as secretariats for political leadership has been enhanced through this process.

The Norwegian reform process has mainly been a combination of internal delegation of authority combined with a more formalized performance assessment regime and external structural devolution. Devolution can take many forms, but one main idea is to strengthen the discretionary power of managers and give subordinate levels of the administration, agencies and public enterprises more autonomy (Christensen and Lægreid 2001b, Grønlie and Selle 1998, OECD 1996). Structural devolution entails a transfer of authority downwards in the hierarchy, either to lower-ranking units within existing organizations or to new subordinate governmental organizations, and implies increased vertical specialization (Christensen 2001). Many governments are moving units or tasks to organizational forms that are vertically further away from the central political leadership. Global pressure to cooperate and compete in new ways is pointing in the direction of more market competition and imperative change processes that a single country or administrative unit cannot easily cope with. This is a reflection also of the doctrine that politics and business should be separated and that private actors are better market actors than public ones. Hence, it is claimed that the most effective way of doing business for the public sector is either to create organizational forms that attend more systematically to commercial functions or to let private actors take over (Boston et al. 1996).

(17)

There are, however, important distinctions between the Norwegian reform programmes and the main trends in the international administrative policy reform movement (Olsen 1996). First, visions of the decline of the public sector through privatization scarcely featured in the Norwegian reform programmes (Christensen and Lægreid 1998a). Second, international criticism that the growth of powerful special interest organizations had prevented political institutions from addressing broader national concerns did not receive much support in Norway.

A mixture of symbols and rhetoric, on the one hand, and pragmatic and instrumental measures, on the other, has characterized the reform programmes in Norway. The slogan of the government of today «From words to action» indicates a wish to weaken the symbolic elements of administrative reform and to enhance the substantive outcome (Christensen and Lægreid 2003a).

The formal structural arrangements in the Norwegian state apparatus

The existing forms of state organizations in Norway are outlined in Table 1, starting by those closest to the executive on the left and increasing degrees of structural devolution towards the right.

(18)

Table 1. Different forms of affiliation for units in the Norwegian state and the number of units in each category. 2003.

Civil Service State-Owned Companies (SOC)

Governmental foundations

Ministries Directo- rates, Central agencies⁴

Other public administration bodies

Central Agencies with extended authority

Govern- ment administrative enterprises

Financial inst., funds

Hybrid companies established by special laws

Govern- ment- owned com-panies

Govern- ment limited companies (100%

state owned)

Limited companies with the state as majority owner

Central foundations

Fringe foundations

18⁵ 62 90 21 5 5 9 6 28 19 52 26

Example:

Ministry of Health

Example:

Railway Inspecto- rate

Example:

Norwegian Polar Institute

Example:

Research Council of Norway

Example:

Norwe- gian Mapping Authority

Example:

Norw.

State Housing Bank

Example:

Retail Monopoly of Wine and Spirits

Example:

Norwe- gian Power Company

Example:

Norwegian State Railways

Example:

Statoil Example:

Museums Example:

Telemark Research

18 183 62 78

The major formal dividing line, as shown in Table 1, runs between central administrative units, including government administrative enterprises, and state-owned companies (Christensen and Lægreid 2003b). There is also a significant difference between governmental foundations and SOCs. Central agencies and government administrative enterprises are units at sub-ministerial level and are, legally speaking, government entities subject to ministerial directions and directly subordinated to ministerial control.

In contrast to state-owned companies, the state budget, the state collective wage agreement, the state pension scheme, the Freedom of Information Act, and the administrative law regulate the civil service.⁶ Government administrative enterprises are given enhanced budgetary leeway.

One can distinguish between internal structural devolution, whereby civil service organizations are converted from one sub-category into another, giving them more autonomy from the political leadership, and external structural devolution, whereby civil service organizations are converted into some sort of state-owned company or governmental foundation (Lægreid and Roness 1998). The main aim of this process is to distance the delivery and control of an activity or service from politicians and to secure commercial benefits and professional autonomy. This is done by transferring some

4 Including regulatory and scrutiny agencies for the SOCs.

5 Including Prime Minister’s office

6 See Appendix for a more detailed description

(19)

formal powers from the minister to a board or to the agency or company leader, where it can be exercised with a degree of autonomy from ministerial control, even though the responsible minister does not relinquish his or her authority entirely. Its ideological basis is a new and narrower view of the role of the state as a «supermarket state», where the primary emphasis is on the state as a regulator of market competition (Olsen 1988) rather than as a comprehensive, sovereign planning state.

The reforms of state-owned companies in Norway during the last decade have brought some substantial changes to the traditional structure. SOCs are primarily characterized by their independent legal status, by having their own control or scrutiny bodies, by holding responsibility for their own economic resources, and closely observing the laws regulating private companies (Zuna 2001). The government controls the SOCs through its ownership position, manifested in an annual general assembly and in some cases a continuous steering dialogue between the owner and the companies, but also through its role of regulator by laws and regulations. The companies' income comes from the market but also from the government, which pays the companies for non- profitable social functions or simply transfers money to strengthen their economic basis.

SOCs can handle their own income independently of economic decisions made in the Storting (Parliament). They can borrow money to finance their activities, and the responsibility of the government for the loans is limited. Government limited companies can, in principle, go bankrupt. The different forms of company affiliation are displayed in Table 1, with the structurally most devolved forms – in the sense of autonomy from guidelines and government steering – in the far right-hand column.

Studies of how the different types of new SOCs work indicate that government control of these companies is seen as overly formal and relatively passive, and also that in practice they have many similar features, with little differentiation between them (Statskonsult 1998). There is no close match between the legal status and operational practice of each sub-category (Grønlie 1998, Wik 2001), which means that for analytical purposes we can treat them as one category of SOCs. A plausible conclusion from the existing studies is that while organizational form matters, it does not closely determine the pattern of control and autonomy. Even if the form of affiliation allows significant leeway for a trade-off between different objectives and values, it is fair to say that, generally speaking, the possibilities for political control decrease and the latitude for economic values and market considerations increases as we move towards the right of Table 1 (Wik 2001). The companies’ contacts with ministries drop when they are converted from government administrative enterprises into SOCs, and less weight than before is given to political signals and non-commercial activities (Christensen and Egeberg 1997).

Governmental foundations are separate legal entities, founded either by a ministry (central foundations) or by an agency (fringe foundations). Like state-owned companies, they are not covered by the civil service rules and regulations like the state budget, financial management regulations, public personnel administration, the Civil Service Act, the Public Administration Act or the Freedom of Information Act. In contrast to state-owned companies they are self-owned entities, and thus have more formal autonomy from the ministry than SOCs. The government can control the foundations by general laws and regulations, by recruiting board members and by formulations of

(20)

statutes. These control devices are, however, weaker and less precise than for the others forms of affiliation.

Changes in the formal structure of the state apparatus

It is important to distinguish between administrative policy programmes, and the actual changes that are being undertaken. In this section we will concentrate on the changes in formal structure that actually have taken place, by focusing on the growth of administrative units and the horizontal and vertical specialization of the state apparatus.

S i z e o f t h e s t a t e

The number of ministries, including the Prime Minister’s Office, in Norway has increased from 13 in 1947 to 18 in 2003, an increase of 38 %. The corresponding increase in the number of internal administrative units in ministries (divisions, sections etc) is 89%. Our focus in the presentation below is not the ministries and their internal administrative units, but affiliated units at one or more arms-lengths from ministries.

The Norwegian State Administration Database (NSA) has a detailed categorization of different forms of affiliation (Rolland and Ågotnes 2003). Further in the presentation we will focus on three main forms of affiliation: civil service (excluding ministries), state-owned companies, and (governmental) central foundations (see Table 1). But we will also to some extent differentiate between the various sub-categories of civil service organizations and state-owned companies.

It may be added that no subsidiary internal units are included, and public administration bodies having corresponding type of tasks are grouped together. This means that regional offices that are reporting to a national body are not counted, and bodies with corresponding type of tasks reporting directly to a ministry are grouped together and counted as one (for example state colleges). In the NSA, grouped bodies only appear in the sub-category «other public administration bodies».

We will distinguish between four periods in the post World War II administrative history in Norway. First, the period from 1947 to 1969, representing the growth of the Norwegian welfare state. Second, the period from 1970 to 1983. The strong democratisation and political decentralization movement connected to the No-to-EEC referendum in 1972 represents the start of the period. The first conservative government in Norway for more than 50 years when the Willoch government came into office in 1981 represents the change to the third period, introducing the first New Public Management ideas of administrative reform. The last period starts in 1990 with the follow-up of the Hermansen commission and the Brundland government that came into office in 1991, and represents a gradually greater scope and intensity of the NPM- reform movement.

(21)

Figure 2 reveals that there was a steady growth in civil service units (outside the ministries) up to the end of the second period. Then we see a significant decline in number of civil service units.

Figure 2 Number of units within main forms of affiliation 1947–2003.

From Figure 3 we see that the main decrease is in the sub-category «other public administration bodies», many of them being grouped bodies with corresponding type of tasks reporting directly to a ministry (e.g. state schools and research bodies). There has been a significant merging of these kinds of organizations into larger bodies over the past 15 years. Figure 3 also reveals that central agencies are not a new organizational form in Norway. Already in 1947 we had close to 50 central agencies. The number of central agencies was relatively stable up to 1960, representing the start of a 30-year steady but slow increase in the number of central agencies. In the last period there has been a weak decrease in the number of central agencies. The number of government administrative enterprises and central agencies with extended authority has been low, but over the past few years there has been an increase in the number of central agencies with extended authority.

0 50 100 150 200 250 300 350

1947 1949 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

Year

N

Civil service State-owned companies Governmental foundations

1947-70 1970-83 1983-90 1990-03

(22)

Figure 3. Number of units within sub-categories of civil service organizations 1947–2003.

In contrast to the number of civil service organizations, the number of state-owned companies and (governmental) central foundations has increased over time, and especially in the last period for SOCs and over the past two periods for foundations (Figure 2). The first wave of government limited companies was in the 1950s (Figure 4).

After a short cutback period, the number increased from 1966 to 1986. Then the numbers of government limited companies decreased somewhat, mostly because of an increase in the number of partly privatized companies. In 1991, a new sub-category of SOCs was introduced: government-owned companies.

.

0 50 100 150 200 250

1948 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Year

N

Central agency

Central agency with extended authority Government administrative enterprise Other public administration bodies Funds and banks

(23)

Figure 4 Number of units within sub-categories of state-owned companies 1947–2003

In Figure 5 and Figure 6, the number of units within the main forms of affiliation is distributed on parent ministry, in 1948 and 2003 respectively. In 1948, the Ministry of Church Affairs and Education and the Ministry of Agriculture topped the list of civil service organizations. Most of the state-owned companies were subordinated the Ministry of Industry and Shipping, but also six other ministries had SOCs, e.g. the Ministry of Defence. Governmental foundations were mainly subordinated the Ministry of Church Affairs and Education and the Ministry of Social Affairs.

In 2003, the number of civil service organizations is highest in what are now the Ministry of Education and Research and the Ministry of Culture and Church Affairs, while the number of civil service organizations under the Ministry of Agriculture has decreased significantly. Figure 6 also shows that only 4 out of 17 ministries, including the Ministry of Defence, do not have SOC’s. The Ministry of Trade and Industry is the only ministry with more SOCs than civil service organizations. Governmental foundations are most common under the Ministry of Education and Research, but there are also quite a few under the Ministry of Culture and Church Affairs and the Ministry of the Environment. In the last ministerial area, there are more (governmental) central foundations than civil service organizations.

0 5 10 15 20 25 30

1947 1949 1951

1953 1955 1957

1959 1961

1963 1965 1967

1969 1971 1973 1975

1977 1979 1981 1983

1985 1987 1989

1991 1993 1995

1997 1999

2001 2003 Year

N

Government limited companies Hybrid companies Government-owned companies

Limited comp. with the state as majority owner

(24)

Figure 5 Number of units within main forms of affiliation in various ministries. 1948.

Year: 1948

0 5 10 15 20 25 30 35 40 45 50

Church A

ffairs and Educatio Agriculture

Social Affairs Industri and S

hipping

Justice and the Po lice

Finan ce

Defence Fisher

ies

Supply and R ebuilding

Tran sport and C

ommunication Trad e

Foreign A ffairs

Ministry of ..

N

Civil service

State-owned companies Government foundation

(25)

Figure 6. Number of units within main forms of affiliation in various ministries. 2003.

O r g a n i z a t i o n a l c h a n g e o f t h e s t a t e

In the NSA, 23 different exactly defined codes are used to describe organisational change. They are divided into three main groups: founding of organizations (6 codes), organizational maintenance (11 codes) and termination of organizations (6 codes).

As seen in Figure 2, the number of units changes significantly from one year to another. These changes are documented through founding and termination change codes in the NSA. When we break the frequencies of units on (main) form of affiliation, we have to include organizational maintenance change codes that document movement of units in and out of the different (main) forms of affiliation.⁷

In Table 2, organisational changes that influence the number of units between periods for our three main forms of affiliation are summarised. We have two types of foundings and terminations: Pure foundings are (by definition) new organizations with no prior history, while foundings based on existing units are organizations that are being

7 The parallel is the simple population model: If you have a population in a country on two different times, the change in numbers is caused by births, deaths, emigration and immigration. The change in number of units within civil service between two years are caused by the number of foundings, terminations, (existing) units that have changed their form of affiliation to civil service (immigration) and civil service organizations that have changed to other forms of affiliation (emigration).

Year: 2003

0 5 10 15 20 25 30

Educat ion and Resear

Culture and Chu rch Affa

Justice and the Poli

Labour /Government Ad

Agriculture Trade and Ind

ustr Healt

h

Loca

l Gov. / Regional D e Financ

e

Children and Family Affa

the E nvironmen

Defence

Transport a nd Comm

unicati Fisheries Social Affairs

Foreign Affair Petroleu

m and Ener

Ministry of ..

N

Civil service

State-owned companies Government foundation

(26)

created on the basis of one or more existing organizations. Similarly, pure terminations are (by definition) organizations where the tasks of the unit are not continued in other units, while termination into existing units are organizations where tasks continue in one ore more existing units. A more detailed presentation and discussion of the organizational change codes are provided in the documentation of the NSA (cf. also Rolland and Ågotnes 2003).

Table 2 Organizational changes influencing the number of units. 1948 – 2003.

Affiliation Period Number

at the start of

period the

Pure found-

ings

Found- ings based on

existing units

Immi-

gration Emi- gration⁸

Pure terminations

Termi- niations

into existing

units

Number at the end of period the

= + + + - - - =

1948–1970 13 3 0 0 0 0 1 15

1970–1983 15 1 3 0 0 0 2 17

1983–1990 17 1 4 0 0 0 5 17

Ministries

1990–2003 17 5 5 0 0 2 7 18

1948–1970 227 119 15 9 15 21 54 280

1970–1983 280 62 16 7 16 7 33 309

1983–1990 309 20 16 4 25 12 29 283

Civil service

1990–2003 283 41 38 4 37 14 132 183

1948–1970 21 9 4 6 3 4 5 28

1970–1983 28 8 1 8 2 0 3 40

1983–1990 40 4 3 2 5 2 2 40

State-owned companies

1990–2003 40 9 25 12 20 6 14 62

1948–1970 7 10 0 1 0 0 1 17

1970–1983 17 9 0 0 0 0 0 26

1983–1990 26 10 0 6 1 0 2 39

Governmental (central) foundations

1990–2003 39 14 1 2 1 1 2 52

8 This category also covers units that have become subsidiary units in a larger civil service organization, and therefore not counted as a separate organization any longer. In the 1990–2003 period, this applies for 17 organizations, explaining the difference between the number of emigrations from civil service in Table 2 and Figure 7.

(27)

Table 2 shows that the main reason for the decrease in the number of civil service organizations (outside the ministries) from 1990 onwards is merger with other civil service organizations identified as terminations into existing units. This represents a horizontal de-specialization process. Table 2 also reveals that behind the relative stable number of ministries at the end of each period, there has been significant turbulence involving different kinds of foundings and terminations of ministries, especially in the 1990s. The growth in the number of state-owned companies in the last period (from 40 to 62) is mainly a result of foundings based on existing units, but there is also quite a few pure foundings and immigration of units from other forms of affiliation. The growth of the number of (governmental) central foundations is mainly due to pure foundings.

V e r t i c a l s p e c i a l i z a t i o n

Historically, the bulk of government administrative enterprises was organizationally close to political and administrative leadership and had little autonomy but direct access to the leadership. They were defined as part of the central public administration, signaling that political control was more important than commercial values.

During the last 15 years, about 50 state organizations have changed their form of affiliation in the direction of more devolved forms, which accounts for new forms of affiliation with more market competition and commercial freedom and less political control. This number includes external structural devolution from civil service to state- owned companies as well as conversions of SOCs from one type into another more devolved form (for example from hybrid companies established by special laws to government limited companies, or from government limited companies to partly privatized limited companies), and internal structural devolution from ministry to central agencies or from central agencies to central agencies with extended authority or government administrative enterprises (Christensen and Lægreid 2001b). Until 1992, major public activities like railways, telecommunication, power, postal services, forestry, grain sales, public broadcasting, public road construction and airport administration were organized as central agencies or government administrative enterprises. Since then, the commercial parts of these enterprises have been corporatized, meaning established as different types of state-owned companies, while the regulatory parts have retained the central agency form. Adding to this, in 2002 all public hospitals changed their form of affiliation from (mainly) public administration bodies at the county level to state health enterprises. As a result, more autonomous regulatory and controlling agencies have been established (Rubecksen 2003). In several areas, the single purpose organization model has replaced the old integrated civil service model where the roles as owner, regulator, controller, purchaser and provider were conducted within the same organization. Still, there has been reluctance to privatize, although Norway has experienced some recent breakthroughs. For Telenor, the national telephone company, the state sold 17% of its shares to private investors in 2000, and Statoil, the large public oil company, is also partly privatized as of 2001. Telenor is an example of a company that has gone all the way from an integrated public enterprise to a partly privatized limited company (2000) with the interim status of a government limited company (1994). According to the

(28)

company statutes, Telenor and Norway Post were still limited by regional considerations after they changed their form of affiliation, and the public broadcasting still has educational and cultural obligations (Zuna 2001).

Figure 7 shows the vertical specialization as seen from the Civil Service for the last period, 1990–2003. Vertical specialization, or autonomization, can be either internal within the public administration, i.e. from ministries to civil service; or external, from civil service to state-owned companies or to governmental foundations. Figure 7 reveals that in this period four units have moved from ministries to civil service, eight units have moved from civil service to state-owned companies and two have moved to governmental foundations. Four civil service units have been privatized, five have moved to the county level and one to the municipal level. There have been no moves in the opposite direction, towards politization or vertical de-specialization.⁹

Figure 7. Vertical specialization. Central node: Civil Service. Number of units 1990 – 2003

9 The total movement out of the civil service in this period is 20. In addition, 17 units have become part of a larger civil service organization (cf. note 7).

Civil service

State-owned

companies Governmental foundations Parliament

Ministry units

Municipalities

Counties

Non-profitt sector

Market 4

1

0 0 0

4

2 0 8

0

0 0

0 5 0

0

Autonomization Politisiation/

De-specialization

Regionalization Privateizations

The Structural Anatomy of the Norwegian State 1947-2003